Do countries learn from their mistakes? Here we consider one example of this question with respect to banking crises using the concept of effective learning. Excessive credit growth is widely considered to be the most important contributor to banking crises. Thus, it is interesting to see whether banking crises are associated with lower rates of credit growth in the future and if so, what are major factors which influence such changes in behavior. This paper offers an investigation of some of the political economy factors that may influence whether crises result in greater discipline over future credit growth. Overall, we found very little average effective learning in stable autocracies and found considerable discipline in stable democracies; nevertheless, we found even larger discipline effects in countries that transitioned to democracy during and in the wake of a banking crisis.